Feminist Media Studies ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/rfms20 Communicating feminist politics? The doubleedged sword of using social media in a feminist organisation Lee Edwards , Fiona Philip & Ysabel Gerrard To cite this article: Lee Edwards , Fiona Philip & Ysabel Gerrard (2020) Communicating feminist politics? The double-edged sword of using social media in a feminist organisation, Feminist Media Studies, 20:5, 605-622, DOI: 10.1080/14680777.2019.1599036 To link to this article: https://doi.org/10.1080/14680777.2019.1599036 Published online: 24 Apr 2019. Submit your article to this journal Article views: 961 View related articles View Crossmark data Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=rfms20 FEMINIST MEDIA STUDIES 2020, VOL. 20, NO. 5, 605–622 https://doi.org/10.1080/14680777.2019.1599036 Communicating feminist politics? The double-edged sword of using social media in a feminist organisation Lee Edwards a , Fiona Philip b and Ysabel Gerrard c a Department of Media and Communications, London School of Economics and Political Science, London, UK; bIndependent Scholar, Leeds, UK; cDepartment of Sociological Studies, University of Sheffield, Sheffield, UK ABSTRACT ARTICLE HISTORY Media coverage of violence against women and girls (VAWG) has increased in recent years, due to high-profile investigations such as the 2012 Jimmy Savile case in the UK, and in response to the #MeToo movement in the USA. Feminist organisations are likely to be asked for comment by the media as a result, but journalistic interest in case details rather than systemic causes of VAWG means that political messages focused on ending VAWG remain difficult to communicate. In contrast, social media is frequently celebrated as a channel through which the politics of feminist organisations can be promoted more directly, bypassing mainstream media agendas. In this article, we present the results of participatory research that explored the tensions inherent in social media use by one UK feminist organisation, Rape Crisis England & Wales (RCEW). The findings challenge the utopian view of social media as a panacea for news media shortcomings. Rather than being unequivocally positive, integrating social media into a feminist organisation’s communication work is a double-edged sword, bringing significant challenges that users must negotiate on a daily basis. Received 8 April 2018 Revised 13 March 2019 Accepted 20 March 2019 KEYWORDS Rape Crisis; social media; feminist politics; sexual violence; VAWG Introduction Historically, feminist organisations have found it difficult to promote their political agenda in mainstream news. However, recent high-profile cases of violence against women and girls (VAWG)—Jimmy Savile in the UK, Harvey Weinstein and Brett Kavanaugh in the USA, and regular revelations of sexism in elite institutions—arguably offer opportunities for feminist organisations working to end VAWG to promote their aims by placing stories or commenting on cases. From a normative perspective the opportunity to provide media comment is positive. Increased visibility should equate to a higher profile, greater awareness and credibility for the organisation’s work (Kirk Hallahan 2010; Brooke McKeever 2013) and, in the medium term, the possibility of an increase in volunteer and financial support. However, research on media coverage of VAWG shows a persistent pattern of stereotyped victim representations and a focus on individual cases rather than long-term patterns of violence and systemic causes (Jenny CONTACT Lee Edwards l.edwards2@lse.ac.uk Department of Media and Communications, London School of Economics and Political Science, Houghton St, London WC2A 2AE, UK © 2019 Informa UK Limited, trading as Taylor & Francis Group 606 L. EDWARDS ET AL. Kitzinger 2004; Lumsden and Heather Morgan 2017). In contrast, social media is often viewed as a tool through which the importance of ending VAWG can be communicated more directly, without having to accommodate media agendas (Dustin Harp, Josh Grimm and Jaime Loke 2017; Hester Baer 2016). In this article, we present findings from a participatory research project, Communicating Feminism, conducted with Rape Crisis England & Wales (RCEW), which challenge this utopian view of social media. Research focusing on feminist adoption of social media has tended to emphasise its use by grassroots activists and individuals, rather than organisations with an existing political and institutional hinterland (see, for example, Jessalynn Keller, Kaitlynn Mendes, and Jessica Ringrose 2018; Fredrika Thelandersson 2014; Sherri Williams 2016). We argue that, for established feminist organisations, the use of social media is shaped by the organisation’s history, identity and purpose as well as the political, legal and funding context on which survival depends. These factors, combined with social media’s capacity for both openness and surveillance, make choices about using digital platforms for communicating political messages more complicated than is often assumed. They compromise the liberatory potential of digital platforms as tools for “alternative” representations of women, and even have the capacity to shut down, rather than facilitate voice. We conclude that, while social media’s utility for movements such as #MeToo is clear (for example, as a means of rapidly extending audiences, circulating messages and promoting collective action), for established feminist organisations that face a greater number of institutional constraints, it is a doubleedged sword for communicating feminist politics in the public sphere. We begin by briefly reviewing existing research on media coverage of VAWG, and the potential social media offers to feminist and third sector organisations. We then introduce RCEW, the project and methodology, and the findings. We conclude by considering the complicated and difficult realities of using social media for feminist organisations. Media coverage of VAWG and the promise of social media Rape and sexual violence began to receive regular coverage in British newspapers from the 1970s onwards, rising throughout the 1980s (Keith Soothill and Sylvia Walby 1991), when child sexual abuse (CSA) was brought to public attention via a spate of scandals (J. Kitzinger 2001; Paula Skidmore 1998). The 1980s also saw the sexualisation, sensationalisation and tabloidisation of news under new commercial pressures that resulted in a shift towards an “infotainment” register (Karen Boyle 2005; Rosalind Gill 2007). These critical feminist analyses of media coverage of sexual violence in Anglophone Western democracies have repeatedly drawn attention to the construction and circulation of falsehoods about the incidence and realities of sexual abuse; the focus on individual, often atypical, cases versus systemic (gendered and intersectional) framings; and, more recently, the sexualisation of coverage. Rape reporting is “de-gendered” as male perpetrators are invisibilised in headlines and official statistics, or blame is apportioned elsewhere (e.g., to date rape drugs) (Boyle 2005; Gill 2007). Studies across other national contexts illustrate the tenacity of rape myths in news coverage, with victims implicated in their attacks, racist stereotypes of perpetrators recurring alongside geopolitical tropes, and coverage trained through a sexualised lens (Zeynep Alat 2006; Meenakshi Gigi Durham 2015; Stephanie Bonnes 2013). FEMINIST MEDIA STUDIES 607 Journalistic structures and practices have contributed to these patterns through a lack of specialist journalists covering CSA stories and the use of “official agencies”— court reports and police—as primary news sources, marginalising feminist campaigners (Skidmore 1998). As a result, misogynist news scripts about sexual violence continue, alongside more subtle post-feminist framings that, to some extent, mainstream feminist messaging (Boyle 2005, 2017; Durham 2015; Lumsden and Morgan 2017; Deb Waterhouse-Watson 2016). The shortcomings of mainstream media for communicating the politics of VAWG, and the increasing importance of digital channels, have resulted in social media playing an increasingly important role in feminist organisations’ communication campaigns. ICT innovations have long been the focus of feminist imaginings, from Donna Haraway’s 1985 “The Cyborg Manifesto” and forecasts of a rebooted (cyber-)feminism in the 1990s (Sadie Plant 1996), to current discussions of movements kick-started online, such as UK Feminista and the Everyday Sexism Project. Andi Zeisler (2013, 179) notes “there is much more of a chance that these [feminist perspectives] will actually have a chance to speak as loudly as the dominant media simply by virtue of being readily accessible.” Feminist movements have used social media alongside framing tactics and direct relationshipbuilding work with journalists, to contest dominant narratives about VAWG, improve coverage of women’s lives and experiences, and mobilise action across constituencies and geographies (Danica Minic 2014; Barbara Barnett 2005; Henrike Knappe and Sabine Lang 2014; Harp, Grimm and Loke 2017; Baer 2016). Social media platforms also provide places where women can develop feminist “counterpublics,” learning about and challenging misogynistic content, rape myths and other justifications for VAWG (Baer 2016; S. Sills, C. Pickens, K. Beach, L. Jones, O. Calder-Dawe, P. Benton-Greig, and N. Gavey 2016; Keller, Mendes, and Ringrose 2018). More generally, social media’s potential has been recognised by organisational communication and public relations scholars. Social media channels outpace the speed and reach of traditional media, are relatively cheap, and can facilitate relationship-building and dialogue with audiences (Michael Kent 2013; Gregory Saxton and Richard Waters 2014). It is now commonplace for organisations, including non-profits, to use social media platforms alongside traditional media in communications strategies (Rowena Briones, B. Kuch, B. F. Liu, and Y. Jin 2011; Gregory Saxton and Chao Guo 2014; Erich Sommerfeldt 2013; Donald Wright and Michelle Hinson 2012). Social media creates spaces for activist communities to develop, generating networks of like-minded individuals and organisations, facilitating conversations that contribute to a common identity and establishing a networked, counter-public sphere for debates (Cheryll Soriano 2014; Thelandersson 2014; Giselle Auger 2014; Aristea Fotopoulou 2016; Williams 2016). They offer the possibility for “activists to enter into news and information flows and overcome journalistic practices that may limit the reporting of their activities to less meaningful frames” (Libby Lester and Brett Hutchins 2012, 848), interrupting the media agenda by increasing the visibility of counter-publics and their narratives (Brian Loader and Dan Mercea 2011). In these “relational sphere[s] of interaction” (I. Himelboim, G. Golan, B. Moon, and R. Suto 2014, 361), audiences, organisations and other stakeholders create connections that influence the outcomes of communication (Saxton and Guo 2014; Hilary Fussell Sisco and Tina McCorkindale 2013; Himelboim et al. 2014; Richard Hanna, Andrew Rohm, and Victoria Crittenden 2011). 608 L. EDWARDS ET AL. These advantages notwithstanding, social media can be time- and resource-intensive and require technical competence, all of which can be in short supply in non-profit organisations (Sommerfeldt 2013; Andreas Schwarz and Alexander Fritsch 2014). Communicating effectively requires organisations to understand the complexity of the social media landscape, the capabilities of different platforms, their use by audiences and their suitability for communicating different types of information (Saxton and Guo 2014; Saxton and Waters 2014; Sisco and McCorkindale 2013). Content must be engaging and offer something audiences can relate to (Auger 2014). In other words, organisations must engage in ongoing connective labour, the “largely invisible digital, immaterial, and affective labor” (Megan Boler, A. Macdonald, C. Nitsou, and A. Harris 2014, 3) that characterises the strategic use of digital media to generate “affective glue” that can hold a movement together. However, the demands of such work mean the promise of dialogue and connectedness often go unfulfilled, with social media used mainly for information-sharing instead (Saxton and Waters 2014; Kent 2013). In summary, research suggests that social media may provide an alternative space for feminist organisations to bypass mainstream media agendas. However, the complexities of running social media may be challenging. While these findings are important, they have been based largely on case studies and quantitative data; the lived experience of using social media in feminist organisations remains under-explored. In a climate of austerity, where such organisations are struggling to survive, collaborative research is increasingly scarce due to time and resource limitations, but is crucial for a fuller understanding of how social media is affecting their work. This article offers a rare snapshot of how feminist organisations may grapple with the communicative opportunities and challenges of social media (see also Zeisler 2013). RCEW and communicating feminism RCEW was established in 2003 as an umbrella organisation for 46 Rape Crisis Centres across England and Wales.1 Rape Crisis Centres are feminist charities offering support for victims and survivors2 of sexual violence, including helplines, face-to-face counselling, advocacy services and, in some cases, text services. They also work towards ending VAWG by raising awareness of its prevalence and effects through communication campaigns demythologising rape and contesting ingrained beliefs about perpetrators and survivors. Their funding is a precarious combination of state support, institutional grants and individual donations. They are asked regularly for input on news stories, but struggle to promote their feminist message and have had limited success driving the news agenda, despite increased media attention to sexual abuse cases. In response, RCEW has considered how to increase its use of social media to more proactively communicate its feminist aims. RCEW launched its first website (http://rapecrisis.org.uk/) in 2004 and began to move away from the “‘head down and get on with things’ approach [that had] led to [its] unintentionally being a fairly insular movement” (Westmarland cited in Helen Jones and Kate Cook 2008, x). In recent years, RCEW has worked more actively with the media, developing a Communications Strategy in 2008 and joining Twitter and Facebook in 2011 (some Centres started using social media earlier than this). Most Centres run websites and some use social media. RCEW appointed a part-time salaried Media and FEMINIST MEDIA STUDIES 609 Communications Coordinator in 2013,3 who facilitated a “Media Tree”—a network of women based in Centres across England and Wales who field media enquiries. Communicating Feminism was a participatory action research project co-designed by the authors Fiona Philp and Lee Edwards, and RCEW’s Media and Communications Coordinator, to explore how social media might be used to proactively communicate RCEW’s political agenda. It was conducted from 2014 to 2016, a period when UK media interest in VAWG remained high and when RCEW was developing its voice online. The primary aim was to develop a proactive online communications strategy for RCEW. Secondarily, our academic purpose was to critically engage with claims about the democratising potential of social media for feminist organisations struggling for visibility in the digital age4. The project had three stages. First, in March 2014, a day-long brainstorm was organised with five members of RCEW’s “Media Tree,” its Media and Communications Coordinator, and [author names omitted]5. The first half focused on discussing existing media engagement and social media practices; the second half was dedicated to developing a social media strategy. The brainstorm generated five hours of recorded material in total. Our collaborators were selected by the Media and Communications Coordinator and drawn from established and new Centres in the South of England (2), Yorkshire (2), the North-East (1) and the Midlands (1). They included experienced social media users alongside those just getting to grips with digital media in a feminist charity context, salaried workers and volunteers. Two were Centre Chief Executives, two were in charge of social media, one was charged with setting up a new social media service, and one was a regular volunteer delivering a range of support services. All had expertise in supporting survivors. It is worth emphasising how difficult it was to bring the six women together for the brainstorm: the current funding climate and drastic under-resourcing means workers and volunteers are severely time-pressured and supporting survivors must take priority. Therefore, working with a larger group of women would have made the project untenable. We had intended to meet again to finalise the co-produced documents, but this proved impossible. Nonetheless, the collaboration we did achieve, which was facilitated by funding that remunerated our collaborators for their time, delivered valuable insights. The second stage of the project was a quantitative content analysis of media coverage of RCEW. The content analysis was not originally planned, but the brainstorm revealed a need to understand current media coverage of VAWG and RCEW, verify patterns identified by our collaborators and establish whether things were changing in light of the Jimmy Savile case, in order to better frame the social media strategy. We analysed news items in UK print and broadcast media published between October 1 2012 and June 1 2014. Our search terms were tightly focused on identifying the presence of RCEW in coverage. They were: Operation Yewtree (the code name for the Savile investigation) and Rape Crisis; Rape Crisis/Rape Crisis England; the name of RCEW’s media spokesperson and rape/Rape Crisis; and the names of individual perpetrators. Coverage was sourced through searches of Box of Broadcasts6, Lexis Nexis, and RCEW’s media archive. The dataset comprised 133 newspaper stories, 6 items of broadcast coverage, 2 pieces of magazine coverage and 9 pieces of online coverage unrelated to existing broadcasters or newspapers. The low number of broadcast items was 610 L. EDWARDS ET AL. inevitable because broadcasters do not publicly archive news items over long periods and access is therefore limited. The final stage of the project was the co-production of three documents based on the brainstorm outcomes: a Social Media strategy and two Social Media Guidance documents, one for RCEW and one for individual Centres7. Drafts of the documents were circulated to collaborators for input, re-drafted and shared again before being finalised and made available to RCEW’s network. We delivered a training session for Centre staff, based on the Guidance, during RCEW’s national conference in November 2016. In adopting a feminist participatory action research methodology, we wanted to reduce the power relations inherent in research and produce knowledge to address gendered forms of injustice (Ramazanoglu with Holland, 2002). Our collaboration was characterised by openness, reciprocity and a desire to create change (S. Reinharz 1992). Our involvement was both personal and professional. Lee Edwards has a personal commitment to social justice and to understanding how strategic communication can bring this about; this drives much of her academic work. During the project, she also encountered VAWG in her family context, which added a highly personal dimension to her participation. Fiona Philip is a long-time friend of the Media and Communications Coordinator and the Rape Crisis movement; as a queer feminist scholar in a media department during Operation Yewtree, developing an impactful research project in response to the unprecedented events felt politically and personally crucial. The six women from the Rape Crisis movement were collaborators rather than participants: their expertise was at the forefront of the brainstorm and we use their real names in this article rather than pseudonyms, as agreed with them and reflecting this collaborative relationship. We provided a general structure but followed their lead for the direction and substance of the conversations. Their insights also led to outputs that were not originally planned (the content analysis and the Guidance documents for the Centres), and their feedback was incorporated into the final documents to ensure they were effective tools. Thus, “action” took precedence over “research” throughout the project (see also Helen Kennedy, Giles Moss, Chris Birchall and Stylianos Moshonas 2015). Data analysis Our approach to data analysis reflected our desire to reduce power differentials in the research process. We recognise and value our collaborators’ expertise and the findings are structured in terms of what they deemed most significant. However, they did not have time to engage with data analysis, so we conducted this work. We listened to the recordings, identifying sections that spoke to the academic debates we had identified in the literature. We transcribed these sections, re-read them and discussed their implications in an iterative process of reflective engagement with the brainstorm content, the academic literature, and our own experience in the project. As the initial purpose of the brainstorm was to produce the social media strategy, the data reflected this: women who were experienced in running social media for their organisations had more to say, featured more frequently in the relevant sections from the brainstorm, and are cited more often in this article. Areas of agreement did emerge in the discussion and we indicate this by referring to collaborators in the plural. Our participatory approach is reflected through the integration of the voices of our collaborators as experts in their FEMINIST MEDIA STUDIES 611 field, alongside and in conversation with the academic literature. We shared the findings and the final article drafts with the Media and Communications Coordinator, and with those collaborators we were able to reach following the conclusion of the project. For the content analysis, codes were derived from academic literature and the key messages in RCEW’s most recent communications strategy. They included story topic, media type, tone (positive, negative, neutral), actors featured, RCEW commentary, topic of RCEW’s commentary, presentation of RCEW (positive, negative, neutral), and presence of RCEW key messages. Reporting rape: UK coverage of VAWG and RCEW Our content analysis showed that feminist organisations such as RCEW were included as sources by the media when covering rape, and that broader discussions about the trends of VAWG as well as survivors’ voices were also featured, if only very occasionally. Overall, however, the patterns identified in previous research continued during the period we investigated: only 18% of the articles engaged in a general discussion of VAWG and only 12% focused on child sexual abuse (CSA), while 50% focused on specific cases. The police and Crown Prosecution Service were the most frequently featured institutions (30% and 14% of coverage respectively). The perpetrator was featured in 24% of news stories, while survivors featured in just 14%. Other relevant institutions, including the Department of Justice and the National Health Service, were largely ignored8 and there was a general failure to engage with institutions that could implement measures to generate long-term social and behavioural change. The focus on institutional failings and historic CSA meant that a gendered analysis of sexual abuse was sidelined and de-contextualised from the systemic problems that give rise to VAWG, allowing its endemic presence to persist (see also Karen Boyle 2017). The media’s tendency to ignore personal and political dimensions of VAWG was reflected in patterns of RCEW commentary. While 44% of the coverage included comment on the details of a particular case, and the same proportion included comment on the broad trend of VAWG, analysis of causes and consequences featured much less frequently (10% and 23% of stories respectively). In general, RCEW struggled to communicate its wider political agenda unless they could connect messages to a case. Thus, the issue of non-reporting of rape by survivors was covered (21% of stories), as was the fact that rape is devastating for victims (19%), not a trivial event (30%), and always the perpetrator’s fault (20%). However, messaging about institutional causes and longerterm consequences of rape were ignored, the only exception being messages about institutionalised discrimination against female survivors (10%). Our collaborators recognised the paradox of media coverage: they observed the way it perpetuates false ideas about sexual abuse and VAWG, but also acknowledged its power as a site for survivor recognition, witnessing, and for contesting rape myths (see also J. Kitzinger 2000). All found working with media outlets demanding, given the pressures of the 24-hour news cycle and news researchers making “ridiculous requests” (Katie) for responses “within the hour” and with very specific requirements. Survivors were understood as “case studies,” with testimonies edited to fit story formats or completely dropped in the final edit. As Katie noted: 612 L. EDWARDS ET AL. a proportion [of interview requests] are what I call ‘rent a survivor’. Which is polite because [journalists] call them case studies. […] they’ll talk to them for four hours and then edit it down to two lines. Or, they go through a process and it is quite empowering, but then it’s used once. […] Journalists rarely considered the personal investment survivors made in speaking to reporters, their hopes for coverage, or the conditions of their consent: reuse of their stories without permission was one example of the impact that this lack of consideration could have. I’ve heard of a survivor switching on the radio two years later, to [. . .] [be] confronted by herself talking about her experience. A frustrating outcome of media engagement was the erasure of RCEW’s identity as an agent of support and change, despite its long history: “Why do we not spring to mind [as other charities do]? When people develop those empathy feelings, towards survivors, why do they not think of us [but, instead, look to children’s charities]?” (Katie). The promise of social media? In this context, social media potentially allowed RCEW to proactively communicate its feminist agenda and engage more directly with its publics. For organisations working to end VAWG, participating in the counter-publics that social media facilitates is an important way of promoting structural and political change. However, it simultaneously puts the organisation, volunteers and service users in danger of becoming targets for the popular misogyny that has become normalised in response to feminist activism (Sarah Banet-Weiser 2018). In this section, we consider how our collaborators reflected on the promise of social media as a channel for engaging with others about feminist politics in the context of their work and history. Our discussions identified six sources of tension, where the advantages of dynamic, highly public social media platforms co-exist with significant risks. Speaking back to power/exposure and surveillance Our collaborators recognised that the networked technology of social media could increase the impact of their efforts to “speak back” to power. They regarded social media as interconnected with traditional media, but also valued its capacity for public persuasion, amplification and reach, and used this to try and influence journalists. Twitter and Facebook allowed them to respond directly and immediately to news stories perpetuating rape myths and circulate counter-narratives, often in dialogue with mainstream media. For instance, Sarah L. rewrote newspaper headlines, editing them to be consistent with both RCEW’s messaging and legal definitions of age of consent. As she explained: “It’s not ‘child porn’, but ‘child abuse’”; a teenager does not “have sex” with an older man, but “is raped.” Hashtags were used for “guerrilla” tactics—for example, by redeploying others’ hashtags on RCEW tweets, so that a challenging or supportive statement became visible on their timeline. However, speaking back is risky because, unlike campaigning-only organisations, RCEW has to manage political and funder sensibilities. Consequently, social media FEMINIST MEDIA STUDIES 613 communication needed to be “thought-through” (Sarah L.). Those with a vested interested in RCEW communication came from many quarters: survivors (see Survivors’ needs/ Survivors’ risks, below), employers, donors, policy makers, and the sector’s regulator, the Charity Commission. Twitter’s capacity to enable constant surveillance of RCEW and Centres led some collaborators to use alternative, non-institutional addresses/personas for messages that were more politically “edgy” than an RCEW-owned account allowed. When organisations mask their identities—for example, in cases of greenwashing or front groups—it can be a cause for concern because it disguises the pursuit of hegemonic power. However, in this case using a different identity was dictated by the aggressive and surveillant social media environment that mitigates against feminist organisations’ activism. When using alternative addresses, our collaborators did not speak as RCEW, but as feminist activists who wanted to make political points, but could not do so under the constraints of their organisational membership. Nor did they wish to introduce controversy that could be linked to RCEW for fear of endangering its survival. Consequently, alternative addresses offered a route for them to exercise personal political resistance while simultaneously protecting the organisation’s future. Their experiences reflect research that reveals online spaces as places where male surveillance of feminist activity is rife, shaping the ways that women can “talk” online (Jessica Megarry 2017), but also pointed to offline political-economic conditions as important limitations for activism. An angry tweet could invite accusations of libel, while being too political ran the risk of Charity Commission sanctions and funding being withdrawn. As Katie noted: “There’s lots of people watching [. . .] I feel like we could be pounced on. […] I drool with envy over what [purely] campaigning groups can do.” Extending reach/controlling outcomes Our collaborators used the network capabilities of social media in a range of ways. They connected with activist groups pursuing similar agendas to access networks that would allow RCEW messages to circulate widely; they made the most of their internal network by re-tweeting other Centre’s tweets; and they gathered ideas about responses to topical issues, adopting what they described as a “dip test” approach of monitoring the environment to see how other activist organisations were commenting on stories. Social media was also useful for sharing knowledge, opinion, and engaging supporters in speaking “for” RCEW—in particular, making politicised comments that would be too sensitive for RCEW to contribute. This allowed different voices to be included in debates about VAWG. Katie noted: “I often go to EVAW [End Violence Against Women Coalition], bloggers, EVB [Everyday Victim Blaming], refugees and asylum seekers women’s organisations, comedians. Also Camilla Parker Bowles, politicians, MumsNet, Imkaan.” Reciprocal relationships extended the reach of RCEW’s communication and supported offline alliances with sympathetic journalists as well as supporters from the celebrity and entertainment worlds. Twitter and Facebook were a potential resource for journalists to pick up stories, encouraging requests for comment, raising RCEW’s profile and developing relations with local media (whose coverage was crucial for ensuring Centre messages reached local survivors). 614 L. EDWARDS ET AL. However, while RCEW benefited from messages that moved rapidly and widely across networks, they also ran the risk of quickly losing control of the desired meaning of those messages. Moreover, communication by other actors could be problematic because the consistency of their commentary could not be guaranteed. Katie explained how one high-profile comedian, who had worked on a successful fundraising initiative, subsequently made problematic jokes about VAWG in her show. Moreover, investing time and resources in alliances did not always bring the expected benefits: one popular website was an enthusiastic partner, but raised only very limited funds for RCEW. Survivors’ needs/survivors’ risks The organisation’s mission to support survivors was keenly felt by all the collaborators (as one participant put it: “Everything we do is to help survivors”) and so the first question to consider for any initiative was “is this going to discourage a survivor [from asking for help]?” They focused on crafting a positive, supportive “voice” for communication, given its potential impact on survivors’ well-being. Sarah L. noted the importance of maintaining hopefulness, commenting: “We discovered that bleak doesn’t work.” The preferred tone was “about being expert, being credible and being up-todate. It’s very calm. We challenge but only through questioning. So, we don’t make any really strong statements, because we’re constantly watched by survivors” (Sarah L.). Communication was also guided by the desire to counter misconceptions about survivors. Yvonne explained: “[T]here is this ingrained belief by the general public that in some way it was a woman’s fault that it happened. And that’s what we need to get rid of.” Two principles were fundamental: believing the victim (in contrast with the judicial system where victimhood may not be recognised until vindicated in court); and facilitating survivors’ voices in order to change attitudes. While these principles meant that social media communication was frequently a positive intervention in discourses about VAWG, its interactive nature presented difficulties in terms of RCEW’s commitment to prevent harm. Positive communication encouraged survivors to respond, but this could lead to problems because the apparently intimate, one-to-one nature of a Twitter feed sometimes prompted women to disclose experiences, which put them at risk of trolling. As recent cases have shown, high-profile feminist activists can be subjected to aggressive and threatening behaviour online, offline and in encounters with journalists (Minic 2014) and the rise of “cybersexism” is well-documented (Laurie Penny 2013; Banet-Weiser 2018). All our collaborators were clear that while dialogue with survivors was desirable, disclosure should be managed by moving conversations offline. Sometimes, problems arose because angry survivors could appear as trolls, and the format of social media makes it impossible to tell initially whether an attack is genuine, personal, or focused on the movement to end VAWG. Managing such challenges was based on a form of “embodied understanding” (Mark Johnson 2015) that engaged both a rational analysis of the situation, and an instinctive “reading” of communication from both survivors and trolls, built up over years of experience and emotional engagement in the sector. Both were essential in the politicised and conflict-ridden context of RCEW’s work. On the one hand, rationality was reflected in Sarah L’s observation: “It’s very easy FEMINIST MEDIA STUDIES 615 to misunderstand 140 characters so you need to check you are interpreting correctly, ask for clarification.” On the other hand, experience gave our collaborators a “sixth sense” about the approach to take. Combining these insights allowed them to identify appropriate options for shutting down trolling, including adopting an evidence-based, expert response using facts and figures, and using “stock” responses that concluded conversations. Communicating on social media also presented more general risks to survivors. Some collaborators tried not to engage with trolls at all, for fear of “gaslighting”— prompting survivors to doubt themselves. As one collaborator put it, “when that stuff is out there it is dangerous”; refusing to extend the conversation was one way of removing it from view. Others noted that raising the public and media profile of VAWG cases and issues necessarily introduced the possibility of triggering survivors’ experiences. Moreover, choosing to feature or comment on one survivor’s story inevitably meant excluding others, which could be perceived as an assessment of worth. As Katie noted: “for every story you choose to include, you don’t choose a different story.” Thus, the possibility of being a source of new survivor trauma coexisted with providing hope and recovery. Facilitating voice/perpetuating silence As feminist organisations, RCEW and the Centres aim to actively challenge power in order to end VAWG; our collaborators envisioned social media as a powerful tool to help them achieve this, a platform for giving voice to survivors and their stories and removing their dependency on traditional media to tell survivor stories publically: “I want survivors to tell their story in the way they want to” (Sarah L.). From this perspective, social media was a way of wresting back representational control from journalists, who did not work sensitively with survivors. It was also a route for unmediated communication between survivors, “liberating” them from the silence they often endure, so that they are able to communicate directly with each other (Keller, Mendes, and Ringrose 2018). Sarah L. described the idea/l of removing the role her Centre played in communicating on behalf of survivors: I‘d like to take [the Centre] out of the middle—survivors talking to survivors—[direct] testimony. A lot of women want to tell their story, they really want to tell their story— have your day in court is a false trope […] there is really powerful testimony. In this vision, social media platforms allow for pluralism, for telling many stories that reflect and connect a multitude of differently situated women and girls. However, facilitating this was not simply a matter of connecting people online or providing access to feminist discussions. As noted, popular feminism has prompted a parallel rise in popular misogyny (Banet-Weiser 2018) and Rape Crisis volunteers had to manage survivors’ contributions to ensure anonymity and protect them from trolls and other unwanted attacks. Providing this kind of support costs money, time and expert labour in terms of editing content to remove identifiers, as well as liaising with survivors to ensure their voices are retained in posted stories (Sarah L.). Thus, maintaining the integrity of service provision, manifest in the relationship between Centres and survivors, meant 616 L. EDWARDS ET AL. that the feminist politics of making survivors visible and heard had to be tempered by the need to accommodate their vulnerabilities. As Sarah L. explained: [e]verything that we do on social media is to try and crowbar open [discussion of VAWG].. [but] because we’re such a confidential service, we mirror that with silencing, [. . .] So the confidentiality and safety has to be there but our job is to rip that open. Saving/taking time and resources Our collaborators noted significant benefits of social media relating to time and resources: the interconnectedness of platforms and availability of social media management tools certainly offered advantages for the speed, reach and ease of their communication. It allowed them to rapidly scale up debates, respond to news in real time, and free up time for dealing with other media enquiries or campaigning. Platforms were used for different styles of communication in order to better tailor messages to audiences. Facebook, for example, was seen as a more “local” platform (Katie), more personal and to some extent more intimate. It was slower than Twitter, but could carry more detailed explanations of media stories and cases. Similarly, websites offered more scope for putting up “meaty,” detailed discussions as well as making basic information (opening times, services provided) permanently available. Nonetheless, our collaborators also encountered problems identified in previous research, including lack of funds and time for digital development in an age of austerity; generational and national differences in digital literacy, competencies and confidence; and the fear of surveillance and governance through networked technologies (Fotopoulou 2016; Helen Thornham and Elke Weissman 2013). For RCEW and the Centres, social media was not always time- or labour-saving; using channels effectively depended on having a dedicated individual, both in the sense that their main task was to manage the social media feeds, and that they were committed to giving up significant time to manage the communication alongside other roles. While one collaborator used a social media management tool, most ran the Twitter accounts manually. For the few people who had both time and an understanding of the technology, but also juggled their own personal pressures, it was a burdensome responsibility. The speed, consistency and reliability of social media communication was always at risk, while slow or inappropriate responses could damage a Centre’s reputation. Yvonne noted the need to make the most of media interest in VAWG and ideally, comment on every story that appeared, but these practical limitations made it extremely difficult. As Sarah L. noted, “I don’t want to start any conversations that I can’t finish in real time, because it sweeps past so quickly.” To some extent, these problems were exacerbated by one of social media’s normative advantages—its extensive reach. Tweets that reached international audiences potentially compromised the local identity and purpose of Centres; while an international reception for their messages was positive, responding to survivors in different countries took up scarce time and resources required for local women. All collaborators commented on the impossibility of stretching resources far enough to make the most of the opportunities that social media offer. FEMINIST MEDIA STUDIES 617 Controlling voice and identity/facilitating diversity Normative organisational communication theory suggests that a singular voice on social media helps develop and support a consistent identity and in the context of activist communication, can frame common causes and risks in a way that constructs “activist imaginaries” that appeal to audiences (Camilla Reestorff 2014, 7; McKeever 2013). At the same time, visibility presents an institutional risk: it can reveal the internal debates and contestations that are part of many activist movements (Thelandersson 2014; Katalin Fábián 2002), fragmenting identities and messages and reducing clarity. Our discussions reflected this tension. Our collaborators recognised that some consistency across the Rape Crisis network was important because it allowed RCEW to more easily manage its multiple roles as campaigning organisation, umbrella body for service providers and lobbyist for women’s rights. Individuals or organisations going “off-piste” could damage RCEW by making it seem too aggressive, or confused in its objectives. Yet, multiplicity was built into the history of the network, and so to create a unified voice for RCEW and all its affiliates was impossible. As our collaborators explained, Centres had their own heritage, location and aims, and their communication needed to reflect a consistent voice and persona across both service delivery and online presence (for Karen, for example, it was “gobby feminist with broader social equality views”). Moreover, as Karen noted, in an organisation supported by volunteers there is less formal control over what people do; those who worked on social media were trusted more often than trained to stick to the organisation’s normative identity and objectives. Training needs themselves were complex, including how to respond to trolls; how to respond to survivors (who might at first sound like trolls); what voice to adopt for a Centre; how to channel anger effectively and the importance of positive messaging. Until the introduction of our Social Media Guidance documents, RCEW was not able to resource such training, so other sources of expertise were often used (for example, material from the Women’s Resource Centre), which were less likely to tailor content to RCEW’s, or its Centres’, identities. Conclusion: the double-edged sword of social media Our project confirmed that the opportunities and limitations of social media identified in previous research certainly applied to RCEW and the Centres, but also revealed how their communication activities were characterised by a balancing act determined by the history and identity of the organisations. Communication using social media was an ambiguous strategy. It acted as a double-edged sword, reflected in our collaborators’ ongoing “internal” dialogue (articulated to the group during our discussions) weighing up advantages and disadvantages. The tensions presented above illustrated their anxiety about digital technologies even as they recognised the promise of social media for their future feminist imaginings, and the imperative to adopt social media in the contemporary communications environment. Their experiences confirm the findings of other research demonstrating how the use of digital technologies for political engagement is “influenced by a dynamic set of feelings and experiences: enthusiasm, uncertainty and fear” (Fotopoulou 2016, 997). Their situation also echoes the realities of “networked feminism” “characterised by 618 L. EDWARDS ET AL. complex connectivity which operates at the intersections of online and offline, and across campaigning activities, feelings and people” (Fotopoulou 2016, 998). While social media has great potential for visibility and voice, it simultaneously increases the vulnerability of those who participate, making the contemporary politics of communicating feminism more complex, rather than easier to manage. Our collaborators illustrate how communication in these contexts cannot rely only on the rational decision-making that predominates in theories of strategic communication and public relations, but is also grounded in embodied understandings of events that incorporate instinctive and emotional readings of communication built up through experience. Social media also had important practical limitations for RCEW as a campaigning organisation, a service provider and a non-profit institution dependent on external funding, which challenge some of the assumptions underpinning organisational communication and public relations scholarship focused on the advantages of social media as a flexible, fast and far-reaching communications channel. For RCEW, even if social media campaigns were successful from a communicative perspective (that is, messages widely shared and circulated, new networks and allies developed), they had only a limited impact on the ongoing struggle for survival. Limited funding, overwhelming demand and inadequate resourcing were not resolved by the speed or reach of social media that enabled greater visibility for RCEW’s political messages. Managing social media took time and resource away from service delivery and sometimes confused the landscape in which support was offered. Moreover, our content analysis suggests that while social media offer platforms for counter-narratives and “speaking-back,” this does not guarantee change in mainstream media agendas where, as a general rule, coverage of VAWG still neglects feminist messaging. As Boyle (2017) has noted, where institutional power is consolidated through a system of news values and norms that ignore feminist perspectives, simply making those perspectives visible in different channels may be insufficient for changing institutional practices. With the exception of individual relationships cultivated with specific journalists or advocates, “reach” may be limited to an “echo chamber” of sympathetic allies, and ongoing work to influence traditional media coverage of VAWG will remain necessary (see, for example Zero Tolerance 2010). Communicating Feminism challenges utopian views of social media as a communications channel that returns control to its users and facilitates a lively and diverse online public sphere. This normative perspective, often perpetuated in communications scholarship, runs the risk of oversimplifying the complexities of pursuing feminist politics in a digitally mediatised world. For RCEW and its Centres, the advantages of visibility and voice are offset by surveillance and silence; the possibilities of speed and reach are countered by time and resource constraints; and the imperative to communicate runs the risk of removing resources from frontline support. Communicating feminism in contemporary society, even in a context where VAWG is a “hot” media topic, remains a tension-filled, contested and difficult activity. Notes 1. The Rape Crisis Federation (1996–2003) preceded RCEW as the movement’s overarching body. Rape Crisis Scotland is RCEW’s sister organisation in Scotland. 2. “Victim” and “survivor” are contested terms; we have chosen to use “survivor.” FEMINIST MEDIA STUDIES 619 3. In 2016 the organisation restructured and this role no longer exists. 4. While we did not work directly with RCEW service users, the project’s aims are survivorcentric and are directed towards ending VAWG. 5. The Ignite funding covered payment and travel expenses for all RCEW collaborators, which allowed them to participate in the project. 6. Run by British Universities Film & Video Council, Box of Broadcasts is a subscription TV and radio service for education. 7. Following the conclusion of the initial project, we engaged in a pilot of the social media strategy in collaboration with Sarah L., who was conducting separate research into the value of publicising Survivors’ Voices. The pilot is not reported here. 8. Operation Yewtree exposed the BBC and NHS hospitals as locations of sexual abuse, and institutional failings were a regular topic in media coverage (Boyle 2017; Greer and McLaughlin, 2013). However, our dataset only covers stories that featured RCEW during the specified time, and consequently did not include the full range of institutions implicated in Operation Yewtree. Acknowledgments Huge thanks to our six collaborators from RCEW, who so generously shared their expertise, enthusiasm and experience with us. We are grateful to the Creative and Cultural Industries Exchange at the University of Leeds, whose Ignite 2014 grant enabled this energising collaboration. Disclosure statement No potential conflict of interest was reported by the authors. Funding This work was supported by the Creative and Cultural Industries Exchange, University of Leeds [Ignite grant]. Notes on contributors Lee Edwards is a critical scholar of public relations, with a particular focus on the circulation of power and inequality in and through the profession. She has published widely on the production of power in and through public relations, and on diversity in the profession. She draws on a wide range of theory in her work, including feminist perspectives of communication, critical race theory, intersectionality and the sociology of occupational fields. E-mail: l.edwards2@lse.ac.uk Fiona Philip is an independent scholar of gender, sexuality and culture, as well as a rare book seller (Quair Books). She has published on the queer politics of Borderline (1930) and copyright and the cultural industries. Alongside feminist and queer theories, she’s engaged by, and in, the role of ICT for opening up heritage collections and for activist work. E-mail: flphilip@googlemail.com Ysabel Gerrard is a Lecturer in Digital Media and Society in the Department of Sociological Studies, University of Sheffield. She is a former Intern at the Social Media Collective, Microsoft Research New England and is the current European Communication and Education Research Association (ECREA) Digital Culture and Communication young scholars’ representative. Ysabel also organises the Data Power Conference. Her research interests include: social media content moderation (specifically pro-eating disorder communities), anonymous secret-telling apps and teenage life, the ethics of researching social media data, and digital research methods. Ysabel 620 L. EDWARDS ET AL. has published her research in First Monday, the Journal of Communication Inquiry and New Media and Society. E-mail: y.gerrard@sheffield.ac.uk ORCID Lee Edwards http://orcid.org/0000-0001-6542-1234 Fiona Philip http://orcid.org/0000-0003-0344-1584 Ysabel Gerrard http://orcid.org/0000-0003-1298-9365 References Alat, Z. 2006. “News Coverage of Violence against Women.” Feminist Media Studies 6 (3): 295–314. doi:10.1080/14680770600802041. Auger, G. 2014. “Rhetorical Framing: Examining the Message Structure of Nonprofit Organizations on Twitter.” International Journal of Nonprofit and Voluntary Sector Marketing 19 (4): 239–249. doi:10.1002/nvsm.v19.4. Baer, H. 2016. “Redoing Feminism: Digital Activism, Body Politics, and Neoliberalism.” Feminist Media Studies 16 (1): 17–34. doi:10.1080/14680777.2015.1093070. Banet-Weiser, S. 2018. Empowered: Popular Feminism and Popular Misogyny. Durham, NC: Duke University Press. Barnett, B. 2005. “Feminists Shaping News: A Framing Analysis of News Releases from the National Organization for Women.” Journal of Public Relations Research 17 (4): 341–362. doi:10.1207/ s1532754xjprr1704_2. Boler, M., A. Macdonald, C. Nitsou, and A. Harris. 2014. “Connective Labor and Social Media: Women’s Roles in the ‘Leaderless’ Occupy Movement.” Convergence: the International Journal of Research into New Media Technologies 20 (4): 438–460. doi:10.1177/1354856514541353. Bonnes, S. 2013. “Gender and Racial Stereotyping in Rape Coverage.” Feminist Media Studies 13: 208–227. doi:10.1080/14680777.2011.623170. Boyle, K. 2005. Media and Violence. London: Sage. Boyle, K. 2017. “Hiding in Plain Sight: Gender, Sexism and Press Coverage of the Jimmy Savile Case.” Journalism Studies. doi:10.1080/1461670X.2017.1282832. Briones, R.L., B. Kuch, B.F. Liu, and Y. Jin. 2011. “Keeping up with the Digital Age: How the American Red Cross Uses Social Media to Build Relationships.” Public Relations Review 37 (1): 37–43. doi:10.1016/j.pubrev.2010.12.006. Durham, M.G. 2015. “Scene of the Crime: News Discourse of Sexual Assault in India and the Geopolitics of Sexual Assault.” Feminist Media Studies 15 (2): 175–191. doi:10.1080/ 14680777.2014.930061. Fábián, K. 2002. “Cacophony of Voices: Interpretations of Feminism and Its Consequences for Political Action among Hungarian Women‘S Groups.” European Journal of Women‘S Studies 9 (3): 269–290. doi:10.1177/1350506802009003375. Fotopoulou, A. 2016. “Digital and Networked by Default? Women’s Organisations and the Social Imaginary of Networked Feminism.” New Media & Society 18 (6): 989–1005. doi:10.1177/ 1461444814552264. Gill, R. 2007. Gender and Media. Cambridge: Polity. Greer, C. 2013. “The Sir Jimmy Savile Scandal: Child Sexual Abuse and Institutional Denial at The Bbc.” Crime Media Culture: an International Journal 9 (3): 243–263. doi:10.1177/ 1741659013513782. Hallahan, K. 2010. “Public Relations Media.” In The Sage Handbook of Public Relations, edited by R. Heath, 623–642. Thousand Oaks, CA: Sage. Hanna, R., A. Rohm, and V. Crittenden. 2011. “We‘Re All Connected: The Power of the Social Media Ecosystem.” Business Horizons 54 (3): 265–273. doi:10.1016/j.bushor.2011.01.007. FEMINIST MEDIA STUDIES 621 Harp, D., J. Grimm, and J. Loke. 2017. “Rape, Storytelling and Social Media: How Twitter Interrupted the News Media’s Ability to Construct Collective Memory.” Feminist Media Studies. doi:10.1080/ 14680777.2017.1373688. Himelboim, I., G. Golan, B. Moon, and R. Suto. 2014. “A Social Networks Approach to Public Relations on Twitter: Social Mediators and Mediated Public Relations.” Journal of Public Relations Research 26: 359–379. doi:10.1080/1062726X.2014.908724. Johnson, M. 2015. “Embodied Understanding.” Frontiers in Psychology 29 (June). https://www. frontiersin.org/articles/10.3389/fpsyg.2015.00875/full Jones, H., and K. Cook. 2008. Rape Crisis: Responding to Sexual Violence. Lyme Regis: Russell House Publishing. Keller, J., K. Mendes, and J. Ringrose. 2018. “Speaking ‘Unspeakable Things’: Documenting Digital Feminist Responses to Rape Culture.” Journal of Gender Studies 27 (1): 22–36. doi:10.1080/ 09589236.2016.1211511. Kennedy, H., G. Moss, C. Birchall, and S. Moshonas. 2015. “Balancing the Potential and Problems of Digital Methods through Action Research: Methodological Reflections.” Information, Communication & Society 18 (2): 172–186. doi:10.1080/1369118X.2014.946434. Kent, M. 2013. “Using Social Media Dialogically: Public Relations Role in Reviving Democracy.” Public Relations Review 39 (4): 337–345. doi:10.1016/j.pubrev.2013.07.024. Kitzinger, J. 2000. “Media Templates: Patterns of Association and the (Re)Construction of Meaning over Time.” Media, Culture & Society 22 (1): 61–84. doi:10.1177/016344300022001004. Kitzinger, J. 2001. “Transformations of Public and Private Knowledge: Audience Research, Feminism and the Experience of Childhood Sexual Abuse.” Feminist Media Studies 1: 91–104. doi:10.1080/14680770120042882. Kitzinger, J. 2004. Framing Abuse: Media Influence and Public Understanding of Sexual Violence against Children. London: Pluto Press. Knappe, H., and S. Lang. 2014. “Between Whisper and Voice: Online Women’s Movement Outreach in the UK and Germany.” European Journal of Women‘s Studies 21 (4): 361–381. doi:10.1177/ 1350506814541643. Lester, L., and B. Hutchins. 2012. “The Power of the Unseen: Environmental Conflict, the Media and Invisibility.” Media, Culture & Society 34 (7): 847–863. doi:10.1177/0163443712452772. Loader, B., and D. Mercea. 2011. “Networking Democracy? Social Media Innovations and Participatory Politics.” Information, Communication & Society 14 (6): 757–769. doi:10.1080/ 1369118X.2011.592648. Lumsden, K., and H. Morgan. 2017. “Media Framing of Trolling and Online Abuse: Silencing Strategies, Symbolic Violence, and Victim Blaming.” Feminist Media Studies 17 (6): 926–940. doi:10.1080/14680777.2017.1316755. McKeever, B. 2013. “From Awareness to Advocacy: Understanding Nonprofit Communication, Participation, and Support.” Journal of Public Relations Research 25 (4): 307–328. doi:10.1080/ 1062726X.2013.806868. Megarry, J. 2017. “Under the Watchful Eyes of Men: Theorising the Implications of Male Surveillance Practices for Feminist Activism on Social Media.” Feminist Media Studies. doi:10.1080/14680777.2017.1387584. Minic, D. 2014. “Feminist Publicist Strategies: Women’s NGOs’ Media Activism and Television Journalism in Serbia and Croatia.” Media, Culture & Society 36 (2): 133–149. doi:10.1177/ 0163443713515736. Penny, L. 2013. Cybersexism: Sex, Gender and Power on the Internet. London: Bloomsbury. Plant, S. 1996. “On the Matrix: Cyberfeminist Simulations.” In Cultures of the Internet, edited by R. Shields, 170–183. London: Sage. Ramazanoglu. C. with J. Holland. 2002. Feminist Methodologies: Challenges and Choices. London: Sage. Reestorff, C. 2014. “Mediatised Affective Activism: The Activist Imaginary and the Topless Body in the Femen Movement.” Convergence: the International Journal of Research into New Media Technologies 20 (4): 478–495. doi:10.1177/1354856514541358. Reinharz, S. 1992. Feminist Methods in Social Research. Oxford & New York: Oxford University Press. 622 L. EDWARDS ET AL. Saxton, G, and L. Guo. 2014. “Online Stakeholder Targeting and the Acquisition of Social Media Capital.” International Journal of Nonprofit and Voluntary Sector Marketing 19 (4): 286–300. doi:10.1002/nvsm.1504. Saxton, G, and R. Waters. 2014. “What Do Stakeholders Like on Facebook? Examining Public Reactions to Nonprofit Oganizations’ Informational, Promotional, and Community-Building Messages.” Journal of Public Relations Research 26 (3): 280–299. doi:10.1080/1062726X.2014.908721. Schwarz, A., and A. Fritsch. 2014. “Communicating on Behalf of Global Civil Society: Management and Coordination of Public Relations in International Nongovernmental Organizations.” Journal of Public Relations Research 26 (2): 161–183. doi:10.1080/1062726X.2013.864242. Sills, S., C. Pickens, K. Beach, L. Jones, O. Calder-Dawe, P. Benton-Greig, and N. Gavey. 2016. “Rape Culture and Social Media: Young Critics and a Feminist Counterpublic.” Feminist Media Studies 16 (6): 935–951. doi:10.1080/14680777.2015.1137962. Sisco, H. F., and T. McCorkindale. 2013. “Communicating “Pink”: An Analysis of the Communication Strategies, Transparency, and Credibility of Breast Cancer Social Media Sites.” International Journal of Nonprofit and Voluntary Sector Marketing 18 (4): 287–301. doi:10.1002/nvsm.1474. Skidmore, P. 1998. “Gender and the Agenda: News Reporting of Child Sexual Abuse.” In News, Gender and Power, edited by C. Carter, G. Branston, and S. Allen, 204–218. Abingdon: Routledge. Sommerfeldt, E. J. 2013. “Online Power Resource Management: Activist Resource Mobilization, Communication Strategy, and Organizational Structure.” Journal of Public Relations Research 25 (4): 347–367. doi:10.1080/1062726X.2013.806871. Soothill, K., and S. Walby. 1991. Sex Crime in the News. London: Routledge. Soriano, C. 2014. “Constructing Collectivity in Diversity: Online Political Mobilization of a National LGBT Political Party.” Media, Culture & Society 36 (1): 20–36. doi:10.1177/0163443713507812. Thelandersson, F. 2014. “A Less Toxic Feminism: Can the Internet Solve the Age-Old Problem of How to Put Intersectionality Theory into Practice?” Feminist Media Studies 14 (3): 527–530. doi:10.1080/14680777.2014.909169. Thornham, H., and E. Weissman. 2013. “Introduction: Renewing/Re-Tooling Feminisms.” In Renewing Feminisms: Radical Narratives, Fantasies and Futures in Media Studies, edited by H. Thornham and E. Weissman, 1–11. London: IB Tauris. Waterhouse-Watson, D. 2016. “News Media on Trial: Towards a Feminist Ethics of Reporting Footballer Sexual Assault Trials.” Feminist Media Studies 16 (6): 952–967. doi:10.1080/ 14680777.2016.1162827. Williams, S. 2016. “#Sayhername: Using Digital Activism to Document Violence against Black Women.” Feminist Media Studies 16 (5): 922–925. doi:10.1080/14680777.2016.1213574. Wright, D., and M. Hinson. 2012. “Examining How Social and Emerging Media Have Been Used in Public Relations between 2006 and 2012: A Longitudinal Analysis.” Public Relations Journal 64 (4): 1–40. Zeisler, A. 2013. “New Media, New Feminism; Evolving Feminist Analysis in Print, on Teh Web and Beyond.” In Renewing Feminisms: Radical Narratives, Fantasies and Futures in Media Studies, edited by H. Thornham and E. Weissman, 178–184. London: IB Tuaris. Zero Tolerance. 2010. “Handle with Care: A Guide to Responsible Media Reporting of Violence against Women.” Available at: http://www.zerotolerance.org.uk/resources/handle-care-mediaguide?destination=node%2F375 Article Perception and Tolerance of Sexual Harassment: An Examination of Feminist Identity, Sexism, and Gender Roles in a Sample of Chinese Working Women Psychology of Women Quarterly 2020, Vol. 44(2) 217–233 © The Author(s) 2020 Article reuse guidelines: ​sagepub.​com/​journals-­​permissions ​DOI: ​10.​1177/​0361​6843​20903683 ​journals.​sagepub.​com/​home/​pwq Xin Shi1 ‍ ‍and Yong Zheng1 Abstract In East Asian culture, where sex is a sensitive subject, many women still have a high recognition threshold when it comes to sexual harassment, as well as a high tolerance for it. Previous research has shown that feminist identity is effective in promoting women’s physical and mental health and buffering against the negative effects of sexual harassment, thus, it is important to clarify the role that feminist identity plays in the perception of sexual harassment. In this study, we examined whether feminist identity is related to the perception and tolerance of sexual harassment and whether feminist identity mediates the relations between sexism, gender roles, and sexual harassment perception and tolerance among Chinese working women. In a survey of 507 participants, we found that active commitment to feminism was positively correlated with women’s perception of sexual harassment, while passive acceptance of traditional gender roles was positively correlated with tolerance of sexual harassment. Mediation analysis showed that active commitment to feminism mediated the relations between sexism, gender roles, and sexual harassment perception, while passive acceptance of traditional gender roles mediated the relations between sexism and femininity with sexual harassment tolerance. We assert that feminist identity has the potential to enable women to be more perceptive and less tolerant of sexual harassment behaviors, and as such, feminist ideology should be incorporated into education for Chinese women. Keywords feminist identity, sexual harassment perception, sexual harassment tolerance, sexism, gender roles, Chinese working women Sexual harassment is considered to be an extreme type of sexism (e.g., MacKinnon, 1979), and is a prevalent issue in societies around the world. In the United States (U.S.), sexual harassment in the workplace is a major social problem, with 13%–31% of men and 40%–75% of women reporting having experienced it (McDonald, 2012). Likewise, this issue should not be underestimated in China (Parish et al., 2006). As China’s urbanization process has accelerated, women, who are the main victims of sexual harassment (Barboza, 2015; Chamberlain et al., 2008; Fridh et al., 2015), have increasingly become represented in schools, workplaces, and public transportation, where they are at greater risk of sexual harassment (Choi et al., 2016). Parish et al. (2006) used a dataset of 3821 participants to analyze the prevalence of sexual harassment in China, finding that 12.5% of all women and 15.1% of urban women have experienced various forms of sexual harassment. In a survey of 6592 Chinese college students, almost 70% of the respondents (75.8% of the female respondents and 36.6% of the male respondents) said they had encountered different types of sexual harassment (Guangzhou Gender and Sexuality Education Center [GSEC], 2017). As part of the global #MeToo movement, Chinese women started a web-­based campaign against sexual harassment in early 2018, and many women began to talk about their experiences of harassment (BBC, 2018; Fan, 2018; Lin & Yang, 2019). However, China’s #MeToo movement has been controversial, and a section of the public believes that the 1 Key Laboratory of Cognition and Personality (Ministry of Education), Southwest University, Chongqing, China Corresponding Author: Yong Zheng, Faculty of Psychology, Southwest University, Chongqing 400715, China. Email: ​zhengy@​swu.​edu.​cn 218 situations these women describe do not constitute sexual harassment, accusing them of ulterior motives (Lin & Yang, 2019). Furthermore, there is a lack of discussion about the definition and perception of sexual harassment in Chinese society and academia, even though these factors may significantly affect victims’ emotional experiences, subsequent coping strategies, the perception of the validity of the allegations, and even court rulings. A study among Chinese, Malaysians, and Indian individuals found that Chinese participants scored significantly lowest on perceptions of sexual harassment (Yee et al., 2015). Therefore, it is crucial to explore the factors that might affect the perception of sexual harassment among Chinese women. The definition of sexual harassment remains vague and open-­ended (Pina et al., 2009), and many scholars consider sexual harassment to be a psychological construct and a subjective personal experience (McDonald, 2012; Topa Cantisano et al., 2008). Although previous studies demonstrate that women are generally more aware and less tolerant of sexual harassment than men (Blumenthal, 1998; McCabe & Hardman, 2005; Rotundo et al., 2001; Russell & Trigg, 2004; Bitton & Shaul, 2013), not all women recognize the same behaviors as sexual harassment (Fairchild, 2009, 2010). For example, some women may identify a violation of their private space as sexual harassment, while others would not (Grossman, 2008). Past research has identified many variables that predict the perception and tolerance of sexual harassment. Certain individual differences, such as feminist identity, sexism, and gender roles, have been linked to how women perceive and judge sexual harassment (e.g., Bhattacharya & Stockdale, 2016; DeSouza et al., 2007; Russell & Trigg, 2004). An important next step in this domain of research is attaining a clearer understanding of the mechanisms that link these constructs together. Hence, we have investigated five variables of interest: (a) feminist identity, (b) sexism, (c) gender roles, (d) perception of sexual harassment, and (e) tolerance of sexual harassment. Together, these variables help explain how gender-­related attitudes affect women’s awareness and tolerance of sexual harassment. Feminist Identity and Sexual Harassment Feminist attitudes are a set of beliefs that promote gender equality in society (Enns, 1997; Williams & Wittig, 1997; Zucker, 2004). Correspondingly, feminist identity is a collective and social identity based on feminist attitudes (Henderson-­King & Stewart, 1994, 1997, 1999). Previous studies have focused on whether self-­identified feminists experience or are aware of more sexual harassment or sexism, and the findings have provided evidence in support of this hypothesis. Feminists report that they experience more sexism and sexual harassment (Ayres et al., 2009; Buschman & Lenart, 1996; Cowan et al., 1992; Fischer et al., 2000; Leaper & Arias, 2011; Zucker, 2004), which is probably because women who have experienced sexual harassment may find that becoming a feminist is a way to protect Psychology of Women Quarterly 44(2) themselves from harassment in the future (Jackson et al., 1996; Jagger, 1983). Yee et al. (2015) found that prior sexual harassment experiences were related to more sensitivity to sexual harassment among women. Feminists, who report more sexism, are more aware of sexism and sexual harassment in society than non-­ feminists (Downing & Roush, 1985; Henderson-­King & Stewart, 1994, Henderson-­King & Zhermer, 2003; Morgan, 1996; Myaskovsky & Wittig, 1997; Szymanski, 2004; Weis et al., 2018; Williams & Wittig, 1997; Zucker, 2004). This suggests that sexual harassment experiences may encourage women to develop their feminist identity and thus enhance their awareness of sexual harassment. In addition, feminist identity has been shown to be related to less acceptance of sexual harassment myths (i.e., attitudes and beliefs that serve to deny and justify sexual harassment; Lonsway et al., 2008) and more use of approach coping strategies (e.g., confronting the source of sexual harassment or seeking social help) to deal with sexism and sexual harassment (Ayres et al., 2009; Leaper & Arias, 2011; Leaper et al., 2013). Thus, we expected that women with higher levels of feminist identity would be more alert to and less tolerant of sexual harassment. Sexism and Sexual Harassment Sexism has also been associated with perceptions and tolerance of sexual harassment. The ambivalent sexism theory proposed by Glick and Fiske (1996) consists of two types of sexist beliefs: hostile sexism (HS) and benevolent sexism (BS). Hostile sexism refers to a hostile and denigrating attitude toward women, where women are seen as inferior and incompetent relative to men; BS refers to a series of related attitudes toward women. People with BS attitudes view women in prejudiced and role-­restrictive ways (e.g., believing that women need men’s help and protection), but they appear emotionally and subjectively positive, and also induce some prosocial behavior (Glick & Fiske, 1996). Previous studies have suggested that people who are more hostilely sexist are less inclined to judge a case as constituting sexual harassment (DeSouza et al., 2007; O'Connor et al., 2004; Ohse & Stockdale, 2008; Wiener & Hurt, 2000; Wiener et al., 1997) and are less supportive of ongoing investigations (DeSouza et al., 2007). Because people who adhere to HS treat women as inferior to men, when sexual harassment occurs, they are more likely to treat harassers’ remarks as innocent and may even interpret these words and deeds as “flirting.” However, the relationship between BS and perception of sexual harassment is less robust, and many studies have failed to find an association between BS and perceptions of sexual harassment (Gutek et al., 1999; O'Connor et al., 2004; Wiener et al., 1997). In addition, to examine the association between sexism and perception of sexual harassment, it is also important to clarify the relation between sexism and tolerance of sexual harassment. People with HS have a higher tolerance for sexual 219 Shi and Zheng harassment or assault, while their gender equality-­oriented peers have a lower tolerance (Chapleau et al., 2007; Hill & Marshall, 2018; Malovich & Stake, 1990; Russell & Trigg, 2004). In addition, hostile attitudes toward women and traditional attitudes toward women were found to be positively correlated with sexual harassment acceptance (Lonsway et al., 2008). It may be that acceptance of sexual harassment is closely related to some aspects of sexism (Lonsway et al., 2008). For example, those who adhere to sexist attitudes are more inclined to objectify and dehumanize women, and this sexual objectification has been shown to be related to acceptance of sexual harassment (Rudman & Mescher, 2012). Individuals who adhere to more hostile sexist attitudes may be more apt to believe in a hierarchy or inherent power differentials when it comes to men and women (i.e., women are inferior to men), or they assume that women may use sexual charisma to gain benefits (e.g., money or a job position) for themselves or to control and destroy harassers (Yamawaki, 2007). Therefore, we hypothesized that both HS and BS would play an important role in determining people’s perception and tolerance of sexual harassment. Gender Roles and Sexual Harassment Gender roles, centered on conceptions of masculinity (usually linked to fortitude and determination) and femininity (usually linked to gentleness and virtuousness), refer to social roles and behavior systems that are identified by certain cultures as suitable for men and women, and also include those basic attitudes and emotions that are traditionally considered to constitute what it means to be male or female (Bem, 1974; Carroll & Wolpe, 1996; Spence & Helmreich, 1978, 1980). The relationship between masculine and feminine gender roles and attitudes toward sexual harassment is not clear, as previous studies have found some equivocal results. Powell (1986) found that highly masculine men were less likely than other men to view offensive sexual remarks as sexual harassment, while highly masculine women were more likely than other women to do so; moreover, both men and women who were highly feminine saw slightly more actions as sexual harassment than did their counterparts. Similarly, Russell and Trigg (2004) found that highly feminine men and women were less likely to tolerate sexual harassment than their less feminine peers, which was in line with the results found by Foulis and McCabe (1997). However, there have been some studies (Bursik, 1992; Bursik & Gefter, 2011; McCabe & Hardman, 2005) that did not find a relation between gender roles and perception and tolerance of sexual harassment. Given that the issue of gender discrimination is more serious in developing countries than in developed countries (Glick et al., 2000, 2004), and that people here adhere more to traditional gender roles, it seems that the differences in the association between gender roles and sexual harassment related attitudes may be more pronounced in the East Asian context. Thus, it is important to examine masculine and feminine gender roles, as an additional gender-­related factor that may be associated with the perception of and tolerance for sexual harassment. Limitations in Past Research There is sufficient research to indicate that feminists are more sensitive to sexual harassment than non-­feminists, but to what extent feminist identity relates to perceptions of sexual harassment is unclear (Bhattacharya & Stockdale, 2016). No studies to date have clearly investigated the relationship between the two. Another important practical consideration is the association between feminist identity and sexual harassment perception from a dynamic perspective. Feminist identity development describes a dynamic process where women move through five developmental stages, from unexamined acceptance of traditional gender roles to development of gender-­equality attitudes to participation in affirmative-­related movements (Downing & Roush, 1985). In previous studies on feminist identity and perceptions of sexism, feminist identity was assessed as a binary construct, namely feminist or non-­ feminist (e.g., Ayres et al., 2009; Buschman & Lenart, 1996; Cowan et al., 1992; Henderson-­ King & Zhermer, 2003; Leaper & Arias, 2011; Morgan, 1996; Robnett & Anderson, 2017; Robnett et al., 2012; Szymanski, 2004; Toller et al., 2004; Weis et al., 2018; Worthen, 2012; Yakushko, 2007). However, a binary approach toward feminist identity risks missing the nuance of how women’s related consciousness and attitudes change during the formation of feminist identity. Admittedly, there are some differences between feminist attitudes and feminist identity (e.g., women who self-­ identify as feminists are more supportive of non-­traditional gender-­roles and participate in more women’s movements than those who embrace some feminist principles but do not self-­ label as feminist; Bay-­ Cheng & Zucker, 2007; Conlin & Heesacker, 2018; Eisele & Stake, 2008; Zucker, 2004). However, many people who support gender equality reject the feminist label (Williams & Wittig, 1997) due to the stigma of the term feminist (e.g., feminists are considered to be offensive and man-­haters). Therefore, the binary approach of using feminist identity may mislabel some women whose feminist consciousness is awakening but who have not yet labeled themselves as “feminist,” thus missing some meaningful findings. Alternatively, despite the fact that the development of feminist identity is seen by many scholars as a dynamic process, only a few studies have been conducted from a non-­binary perspective (e.g., Carpenter & Johnson, 2001; DeBlaere et al., 2017; Erchull et al., 2009; Kucharska, 2018; Sabik & Tylka, 2006; Saunders & Kashubeck-­West, 2006; Zucker, 2004). Furthermore, the vast majority of published studies on this subject were conducted among Western people and college students (Anderson, 2012; Herrera, Herrera, & Expósito, 2014, 2017; Kucharska, 2018; Leaper & Arias, 2011; Robnett et al., 2012; Weis et al., 2018); there has been limited empirical analysis of feminist identity and perceptions of sexual 220 Psychology of Women Quarterly 44(2) harassment in Asia, and none in China specifically. In a cross-­cultural study exploring feminist beliefs and identity among different ethnic groups, Robnett and Anderson (2017) found that Asian women were less likely than European women to identify themselves as feminists. None of the Asian feminists surveyed expressed that they identified as feminists because they endorsed non-­ traditional gender roles. On the other hand, Asian women who did not identify as feminists were more likely to avoid the term precisely because of their endorsement of traditional gender roles (Robnett & Anderson, 2017). This indicates that Asian women, whether feminist or non-­feminist-­identified, appear to be more traditional than Western women in their endorsement of gender-­role attitudes. Therefore, exploring feminist identity and its relation to perceptions of and tolerance for sexual harassment among a sample of Chinese women is important. The second goal of our study was to examine the mediating role of feminist identity in the links between sexism and gender roles (i.e., masculinity and femininity) and perception and tolerance of sexual harassment. Each of the main factors has previously been explored separately in connection with perceptions and tolerance of sexual harassment, but an examination of the relative associations between these variables has not been undertaken in the same study. We hypothesized that a high level of feminist identity (i.e., Synthesis and Active Commitment) would mediate the relationships between (a) sexism and the perception of sexual harassment and (b) gender roles and the perception of sexual harassment (Hypothesis 2a). In addition, we predicted that a low level of feminist identity (i.e., Passive Acceptance) would mediate the relationship between sexism and tolerance of sexual harassment and gender roles and tolerance of sexual harassment (Hypothesis 2b). The Current Study Methods Our study had two related goals. The primary goal was to examine the relations between feminist identity development and perceptions of and tolerance for sexual harassment. For feminist identity, we assessed dimensions of feminist identity attitudes that correspond to Downing and Roush’s (1985) theoretical conceptualization of feminist identity to avoid a binary classification. The first of these stages, Passive Acceptance, is when women recognize and accept traditional gender and social roles, believe men are superior to women, and deny the existence of gender discrimination. Revelation, the second stage, is characterized by anger at men and society and the beginning of questioning traditional gender roles. The third stage is Embeddedness/ Emanation, in which women form an emotional attachment to their own reference group, that is, other women, trying to separate themselves from the male world in order to create an independent space. In Synthesis, the fourth stage, women’s self-­awareness is integrated with their identity outside of gender or group belonging, allowing them to make independent choices based on individual rather than group values. The final stage, Active Commitment, reflects women’s commitment to action in terms of social change with the aim of creating a more equal world. Since women in the Revelation and Embeddedness/Emanation stages of feminist identity have just started to rethink the relations between men and women, women’s identity, and the female community, we would not include the scores of these two stages in our model, only for preliminary analysis. We hypothesized that higher scores on later stages of feminist identity (i.e., Synthesis and Active Commitment) would correspond with higher levels of sexual harassment perception (Hypothesis 1a). We also hypothesized that higher scores on the early stage of feminist identity (i.e., Passive Acceptance) would be significantly related to more tolerance of sexual harassment (Hypothesis 1b). Procedure and Participants Data were collected from December 2017 to February 2018 via an online questionnaire survey posted through the Chinese survey website Wenjuanxing (​www.​sojump.​com), which prevents respondents from filling in the questionnaire more than once based on IP address and collects broad demographic data about the users. This allowed us to recruit participants by age, gender, and profession without a pre-­screening survey. The survey link was widely distributed in a number of discussion groups and chat rooms via some Chinese social media platforms such as QQ, Weibo, Douban, and WeChat. Before the formal part of the questionnaire, an informed consent statement was provided to respondents stating that their demographic data would be collected anonymously and used for academic research only, and that they would be paid 15 CYN (about $2.14) after completing the questionnaire. In order to control for a sequence effect, the measures were randomly ordered in the questionnaire. Participants were eligible for this study if they were female, aged 18 years or older, and employed. At the beginning of the data collection, a total of 13 participants were removed, six of whom were younger than 18 years old, and seven of whom did not fill out more than 20% of their questionnaire. In order to solve this problem, we immediately optimized the settings in the background of the questionnaire survey page. The questionnaires would not be submitted successfully when any of the following conditions occurred: (a) the age of the participants was less than 18 years; (b) participants did not correctly complete the two basic quality control items; or (c) participants did not complete all of the questions. We eventually recruited a sample of 507 heterosexual working women (M = 32.13 years old, SD = 6.52, range = 21–69) from more than 30 cities in China, including 14% (n = 72) from rural areas, 17% (n = 87) from counties, Shi and Zheng and a majority (69%, n = 348) from cities. In the final sample, 3% (n = 14) had no more than high school or vocational school education, 87% (n = 444) had a bachelor’s degree, and approximately 10% (n = 49) had attended some graduate school or received an advanced degree. Measures Feminist identity development. We assessed feminist identity development using the Feminist Identity Composite (FIC; Fischer et al., 2000), which integrated and improved upon two earlier measures of feminist identity development: the Feminist Identity Scale (FIS; Rickard, 1989) and the Feminist Identity Development Scale (FIDS; Bargad & Hyde, 1991). The FIC is a common measure of feminist identity, having been used in more than 20 empirical studies in recent years to explore the relations between feminist identity and other psychological indices among women (DeBlaere et al., 2017). It comprises a total of 33 items, 20 from the FIDS and 13 from the FIS. The FIC consists of five subscales that correspond to Downing and Roush’s (1985) 5-­stage feminist identity model: (a) Passive Acceptance (7 items; for example, “I think that most women will feel most fulfilled by being a wife and a mother”); (b) Revelation (8 items; for example, “My female friends are like me in that we are all angry at men and the ways we have been treated as women”); (c) Embeddedness/Emanation (4 items; for example, “I am very interested in women’s studies”); (d) Synthesis (5 items; for example, “I feel like I have blended my female attributes with my unique personal qualities”); and (e) Active Commitment (9 items; for example, “It is very satisfying to me to be able to use my talents and skills in my work in the women’s movement”). In order to more effectively clarify the associations between various stages of feminist identity and other main factors, we used these subscales as separate scales in this study. Next, we developed a Chinese version of the FIC by inviting two bilingual professionals to translate all items into Chinese. Another person then translated these back into English, and a professional compared the two English versions for equivalence. The word “state” was replaced with “province” in the item “If I were married to a man and my husband was offered a job in another state, it would be my obligation to move in support of his career” based on the administrative region division in China. All items were rated on a 5-­point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). We averaged the item responses within subscales, and high scores for each subscale indicated greater agreement with the corresponding feminist identity stage. The FIC has demonstrated strong convergent validity, and Cronbach’s alphas internal reliability for the scores on subscales were acceptable (DeBlaere et al., 2017; Kucharska, 2018). Reliability coefficients for scores on the subscales in this study were satisfactory, with α = .70 for Passive Acceptance, α = .73 for Revelation, α = .80 for Embeddedness/ 221 Emanation, α = .75 for Synthesis, and α = .77 for Active Commitment. Sexism. The Ambivalent Sexism Inventory (ASI; Glick & Fiske, 1996) was adopted in our study to measure sexist attitudes. It includes two subscales. The first, HS, contains 11 items and is characterized by hostile attitudes toward women (e.g., “Most women fail to appreciate fully all that men do for them”; “Once a woman gets a man to commit to her, she usually tries to put him on a tight leash”). The second, BS, contains 11 items and is characterized by a series of paternalistic attitudes toward women and reflect positive views of women who conform to stereotypical and restrictive gender roles (e.g., “No matter how accomplished he is, a man is not truly complete as a person unless he has the love of a woman”; “Many women have a quality of purity that few men possess”). Each item is rated on a 6-­point Likert-­ type scale ranging from 1 (strongly disagree) to 6 (strongly agree), with higher scores reflecting higher levels of sexist attitudes. The Chinese version of the ASI was developed by Chen and validity support has been documented (Chen et al., 2009; Lee et al., 2010); this scale has also been used in previous studies (Zheng & Zheng, 2015; Zheng et al., 2017). Internal reliabilities for scores on the HS and BS subscales in our study were .78, and .70, respectively. Gender roles. We adopted the Bem Sex Role Inventory (BSRI; Bem, 1974) to assess participants’ gender roles. The first scale to independently measure gender roles, the BSRI was developed by Bem (1974) and includes 60 items, with each item being an adjective describing character traits rated on a 7-­point Likert-­type scale ranging from 1 (never or almost never true) to 7 (always or almost always true). The original scale comprised three subscales: masculine (e.g., assertive, competitive, and willing to take risks), feminine (e.g., cheerful, sensitive to the needs of others, and shy), and neutral (e.g., conscientious, truthful, and happy), with 20 items in each. The Chinese version of the BSRI was developed by Lu and Su (2003), improving reliability and validity of the scale. The new, shorter scale comprises 26 items in two subscales, masculine (14 items) and feminine (12 items). Internal consistency reliabilities for scores on the two subscales were .87 and .78, respectively. In our study, the Cronbach’s alphas for scores on the masculine (α = .84) and feminine (α = .83) subscales were also acceptable. Perception of sexual harassment. Perception of sexual harassment was assessed with 10 items (Popovich et al., 1986), with each item describing a form of sexual harassment. The items were “Eyes me up and down,” “Makes sexual remarks,” “Tells sexual jokes,” “Kisses me on the cheek,” “Asked me to have sex,” “Touches me on arm/ back,” “Asks me on dates after I have refused,” “Touches me on chest/thigh/buttocks,” “Treats me as sex object,” and “Comments on my physical appearance/attractiveness.” 222 Each item is rated on a 5-­point Likert-­type scale from 1 (definitely not sexual harassment) to 5 (definitely sexual harassment). Higher scores indicated more perception of sexual harassment. The scale was originally developed to assess perceptions of sexual harassment among U.S. undergraduate students. A study carried out in Malaysia successfully adopted the scale (Yee et al., 2015) and indicated that the measurement is also applicable to participants in the Asian cultural environment. In the current study, the original version of the scale was translated into Chinese and back-­ translated for validation. Cronbach’s alpha was acceptable in our study (α = .72). Tolerance of sexual harassment. The Sexual Harassment Attitude Scale (Mazer & Percival, 1989) contains 19 items assessing tolerance for sexual harassment and acceptance of sexual harassment myths. Examples of items include “It’s only natural for a woman to use her sexuality as a way of getting ahead in school or at work” and “An attractive woman has to expect sexual advances and should learn how to handle them.” The original version of the scale was also translated into Chinese and back-­translated for validation. Participants were asked to rate items on a 5-­point Likert scale from 1 (strongly disagree) to 5 (strongly agree). Higher scores reflect greater tolerance of sexual harassment. This scale has demonstrated strong concurrent validity, and Cronbach’s alpha internal reliability for the scores on the scale was .84 (Mazer & Percival, 1989). Cronbach’s alpha for scores on this scale was .76 in the present study. Results We used AMOS 21.0 to perform a confirmatory factor analysis (CFA) on the Chinese version of the FIC. The scale’s goodness-­of-­fit indices were χ2/df = 1.59, comparative fit index (CFI) = .92, Tucker Lewis index (TLI) = .90, standardized root mean square residual (SRMR) = .05, root mean square error of approximation (RMSEA) = .03, 90% confidence interval (CI) [.03, .04]. All of the items loaded significantly on their assigned factors (p < .001), and factor loadings ranged from .44 to .90. Consistent with previous research using Western samples and sexual minority women (DeBlaere et al., 2017; Kucharska, 2018), our results support the structural validity of the FIC with a Chinese female sample. Similarly, we also performed a CFA on the Chinese version of the Perception of Sexual Harassment scale. Results supported structural validity, χ2/df = 1.84, p = .011, CFI = .97, TLI = .95, SRMR = .02, RMSEA = .04, 90% CI [.02, .06]. All items loaded significantly on the factor (p < .001), and factor loadings ranged from .43 to .62. Because all of our scales were self-­report scales, we performed Harman’s one-­factor test to check if there was a Psychology of Women Quarterly 44(2) common method bias. The results showed that the first factor accounted for 9% of the variance, indicating that common method bias was not an issue in this study. Preliminary Analyses We used SPSS version 22.0 to conduct an analysis of variance to assess whether there were significant differences in scores for each feminist identity subscale among participants based on age, birthplace, and level of education. The results are presented in Table 1. Only Passive Acceptance, F(3, 503) = 6.112, p < .001, partial η² =.04, and Active Commitment, F(3, 503) = 2.900, p = .035, partial η² =.02, scores differed significantly according to participants’ age. The results of a Least Significant Difference multiple comparisons test showed that participants aged 25 years and younger scored significantly lower on Passive Acceptance, while participants aged 26–35 years were significantly higher on Active Commitment than those in other age groups. In terms of birthplace, Embeddedness/Emanation, Synthesis, and Active Commitment scores differed significantly, Embeddedness/Emanation: F(2, 504) = 5.735, p = .003, partial η² =.02; Synthesis: F(2, 504) = 7.624, p = .001, partial η² =.03; Active Commitment: F(2, 504) = 5.643, p = .004, partial η² =.02. Least Significant Difference test revealed that the scores for these scales were significantly higher among participants from cities than those from counties and rural areas. With regard to education, participants with college, undergraduate, postgraduate, or higher education levels had higher Embeddedness/Emanation, F(3, 503) = 3.765, p = .024, partial η² =.02, and Synthesis, F(3, 503) = 3.462, p = .032, partial η² =.01, scores than those with high school or a lower level of education. However, the data for Active Commitment were not significant, F(3, 503) = .254, p = .776. Table 2 presents the correlations among all the main variables in this study. As expected, the following factors were related to certain stages of feminist identity: perception of sexual harassment, tolerance of sexual harassment, HS, BS, masculinity, and femininity. Feminist Identity and Perceptions and Tolerance of Sexual Harassment To examine the relations between feminist identity development and perceptions and tolerance of sexual harassment, we conducted two hierarchical regression analyses. We decided a priori that we would include demographic variables that were related to FIC subscales as covariates. Thus, we entered age (using continuous data of age), birthplace (using three classifications: rural areas, counties, and cities), and education level (using three classifications: no more than high school, bachelor’s degree, and master’s degree or higher) in step one. In step two, we entered the five FIC subscales. We also performed collinearity diagnostics and results indicated that all predictor variables had acceptable tolerance (tolerance = .63–.96) and 223 Shi and Zheng Table 1. Distribution of Feminist Identity in Demographic Variables. Age [M (SD)]  ≤25  26–35  36–50  ≥50   F value   p value Birth place  Rural areas  County  City   F value   p value Highest education  ≤High school  Bachelor’s degree  ≥Master’s degree   F value   p value PA REV EE SYN AC 20.54(4.46) 22.86(4.93) 23.79(4.12) 23.78(3.11) 6.112
Purchase answer to see full attachment